Strange Problem with Victron Multiplus and LiFePO4

[Dream_Seeker]

Hi Everybody,
After haunting this forum for 15 years, this is my second post!
I’ve owned three Victron Multiplus 12/3000/120 inverter/chargers since 2003 and I have a lot of experience using them with regular flooded and AGM batteries. I even worked with Victron back in 2004 to help them test and debug the Multiplus PowerAssist function to work with the Honda EU-2000 (which ended up working great and Victron subsequently advertised it extensively).
However, I recently encountered a very strange problem with a Multiplus after upgrading my Lagoon cat from 840AH of Lifeline AGMs to a single 460AH Li Time LiFePO4 battery. The upgrade required a lot of changes, including two new Renogy DC-DC chargers, a Victron BlueSolar100/50 charge controller, and a lot of rewiring. The Renogy and BlueSolar chargers worked perfectly all the way to float so I figured a few minutes with VE Configure changing to LiFePO4 and Li Time’s recommended settings would have my 10-year-old Multiplus also working fine but that was not the case.
The bulk phase worked well but only charged at 75-80 amps instead of the specified 92 amps (.2C). However, the moment it reached absorption voltage, it went crazy. Instead of holding constant voltage with current gradually reducing to tail current (9.2 amps) as expected, the voltage varied dramatically – spiking to the high 14 volt range and then dropping into the 13s over about 5 minutes. The current was even more bizarre – quickly dropping to zero amps, slowly climbing back to 55-60 amps, and then quickly back to zero in the same 5-minute cycle. Both patterns repeated indefinitely and the unit never reached float. I wondered if Li Time’s recommended absorption voltage (14.6 volts) was too high and causing a BMS disconnect so I changed to Victron’s default absorption voltage (14.2 volts) but that made no difference. I finally came to the conclusion that my old Multiplus was too old to work with LiFePO4 and I needed an upgrade.
When I looked into buying my fourth Multiplus, I discovered that Inverters-R-Us sold them through Amazon for the same price as everybody else plus their “experts” would pre-configure it and provide “expert” technical support so I ordered from them. Fast forward a week and the new Multiplus was installed and operational. Unfortunately, it displays the EXACT same symptoms in absorption! I called Inverters-R-Us’ “Expert” Tech Support and soon found they knew less about the Multiplus than I do. After carefully preserving their “expert” settings file, I reviewed the VE Configure options and found there are very few changes in the last ten years. The only change I made was to accept Victron’s default LiFePO4 voltages. I also installed a VE.Bus Smart Dongle to see if I could get any insight from Victron Connect but learned nothing new. At this point, I’m stumped!
BTW, the Li Time 460AH battery works GREAT and has more usable capacity than my previous 840AH of Lifeline AGMs did when they were new. As I said before, the other chargers, including the Victron BlueSolar, work perfectly - only the Multiplus has issues. Has anybody else experienced these symptoms with the Multiplus working with LiFePO4?
Thanks!
Glen

[Statistical]

Do you have any other charging sources? If you have solar you may wish to temporarily shut it down. Note never disconnect solar power controller while charging on the battery side, disconnect it on the PV side or put towel/blanket over PV panels. If nothing else this simplifies and isolates the issue.

When the charging is in absorption and "going crazy" what is the voltage at the battery terminals. Use a multimeter. Don't go off the multiplus results.

Lastly why do you need DC to DC converters.

[SailingHarmonie]

I have installed and worked on Li systems with MultiPlus inverter chargers installed. None have the issue you describe. So it’s your system, or its setup details.

Have you tried the Victron Community support pages? Lots of people there with detailed knowledge of the setup conditions that might cause this. Without a full listing of all the setup parameters, I can’t guess what might be wrong.

[smac999]

my best guess is the bms is opening to give those big spikes. but that is not normal for the old or new multi. the new one gives you the ability to tick off lithuim batteries, which removes the battery temp sensor compension, and changes the rebulk setting. that's really all it does. if your old one doesn't have the lith tick box then you need to disconnect the battery temp sensor. but that's not your issue.

try turning down the bulk / absorb to 13.9 and try again. I've had a few drop in batteries that would not charger over 13.8 untill a few days of low charging to balance them. they would shut off. though if the solar is holding them at 14.2 then it's hard to say. the solar probaly doesn't have the power to spike everything in the battery opens though.

[wholybee]

What do you have absorption voltage set at? Almost certainly, the BMS is going open, probably because of a cell imbalance and thus a cell is overvoltage. When the BMS opens, the voltage is going to jump up momentarily to very high, as the current drops to zero, then the Multiplus is going to get confused and all sorts of weirdness.

The challenge here is that you have no way to look inside the BMS and know what exactly is happening. But, the fix is to try and get the BMS to balance the cells (or if new contact LiTime and have them send you a new battery under warranty that is better balanced)

To balance the cells, you want to lower the absorption voltage to something that is low enough the battery doesn't flip out. It needs to be at least 13.8V, otherwise you really do need to contact LtTime. But the higher it is the shorter time it will take. Then, set the absorption time really long- like 2 days. Then, raise the absorption voltage one or two tenths of a volt. You probably don't need 2 days this time. Keep doing that until you get the voltage up to 14.6V.

For day to day use, I would set the Absorption voltage no higher than 14V.

[Svzamae]

I had similar symptoms with my solar controller when I installed my lithium batteries. Turned out to be a bad earth between the controller and the batteries!! I ran a heavier cable outside the loom to test and that fixed the problem. Installed the new cable and now have 30% more input and steady voltage.

[MV Intrigue]

The following is just a theory of one possible reason. Try lowering the bulk/absorb voltage to 14.0 and for the first few cycles lower the charge current a bit. Does this battery have bluetooth? Would be nice to see any warning to see if it was cell overvolt or pack overvolt. Litime, on their forums, have described their "full charge protection" and even though they may list 14.6 as a recommended charge voltage the full charge protection will engage when the voltage reaches any voltage over a certain value (usually much less than 14.6) with current dropping low for a particular time period. This will result is the charge mosfets opening and other odd behaviors such as momentary reduced voltage at the terminals.

On some particular brands of batteries we have learned this threshold might look like this: 14.0 volts or greater > current less than 3.6 amps > for more than 10 seconds and the battery will go into "full chg Prot". Keep in mind it can be charged to higher than 14.0 volts as long as amps dont fall below the threshold, which they eventually will once the voltage threshold is made and the batt gets full at whatever CV is set to.

So if you could find out what those parameters are, you just set your charge voltage just lower so you never allow that voltage threshold to be met. You can find out manually by setting the voltage low, like 13.8, letting amps fall to near 0, then increasing voltage up .1 volts. At that time amps again will increase again and then slowly fall near 0. Then increase another .1 volts and let amps fall again. Do this slowly and you will find your voltage threshold for full charge protection. Then program charge voltage just below this point. This allows your charger and the parameters set into the charger to do the controlling instead of the BMS hard protection limits.

[TheoWjr]

I have seen the multiplus not charging at full capacity when one of the checkboxes on the Charge tab was selected, I believe it was the 'Stop after excessive bulk' but it may have been 'Weak AC input'. I would suggest ensuring that neither is checked a least as a test.

The second issue with amps is quite odd. Do you have any other battery monitor like a smart shunt? Perhaps try selecting or deselecting 'Enable battery monitor' on the General tab of VE configure as a test.
Also are you using any Assistants such as the Two Signal BMS support in your setup?

Cheers,
ted

[CaptainRivet]

Originally Posted by wholybee

What do you have absorption voltage set at? Almost certainly, the BMS is going open, probably because of a cell imbalance and thus a cell is overvoltage. When the BMS opens, the voltage is going to jump up momentarily to very high, as the current drops to zero, then the Multiplus is going to get confused and all sorts of weirdness.

The challenge here is that you have no way to look inside the BMS and know what exactly is happening. But, the fix is to try and get the BMS to balance the cells (or if new contact LiTime and have them send you a new battery under warranty that is better balanced)

To balance the cells, you want to lower the absorption voltage to something that is low enough the battery doesn't flip out. It needs to be at least 13.8V, otherwise you really do need to contact LtTime. But the higher it is the shorter time it will take. Then, set the absorption time really long- like 2 days. Then, raise the absorption voltage one or two tenths of a volt. You probably don't need 2 days this time. Keep doing that until you get the voltage up to 14.6V.

For day to day use, I would set the Absorption voltage no higher than 14V.

I agree with him, most likley an imbalance and BMS shuts off and after 5min the voltage drops during resting under the threshold. The multi with its high current let it spike much faster then the solar controller.
The reduced charging current is most like the AC weak current box ticked in the parameter
To fix i agree to his procedure too, the way to go.